Accuracy of Transrectal Ultrasonography in Preoperative Staging of Rectal Cancer

Abstract

Objective: To determine the diagnostic accuracy. sensitivity, specificity, positive predictive value and negative predictive value of transrectal ultrasound in the detection of the depth of tumor invasion. perirectal fat invasion and regional lymph node involvement in rectal cancer patients at Ramathibodi Hospital.

Materials and Methods: A total of 17 patients with biopsy proven rectal cancer, who had undergone transrectal ultrasound evaluation of the invasion of the rectal wall and the mesorectal lymph nodes status were retrospectively studied during a period of 5 years and 6 months. We compared the Transrectal Ultrasonography (TRUS) staging with pathology examination of the resected specimens according to TNM Classification.

Results: 15 patients had a radical resection (11 abdominoperineal resection and 4 low anterior resection), and two patients had a local transanal excision. Eight among these 17 patients had preoperative chemoradiation. The overall accuracy in assessing the depth of tumor invasion was 47% and 56%. with 41% and 22% of the tumors overstaged and 12% and 22% understaged. in the all patients group and the other group with no neoadjuvant therapy. respectively. In determining the perirectal fat invasion. the diagnostic accuracy of TRUS was 52.9% with sensitivity 83.3% (95% Cl 65.6-101.1). specificity 36.4% (95% Cl 13.5-59.2), PPV 41.7% (95% Cl 18.2-65.1) and NPV 80% (95% Cl 60.9-99.0). The accuracy of TRUS in determining the perirectal fat invasion was higher in the group that excluded neoadjuvant therapy patients at 66.7%, with sensitivity of 66.7% (95% Cl 35.9-94.5), specificity 66.7% (95% Cl 35.9-97.5), PPV 50.0% (95% Cl 17.3-82.7) and NPV 80% (95% Cl 53.9-106.1). The accuracy in assessing nodal involvement in 15 patients treated with radical surgery was 60%. with sensitivity 33.3% (95% Cl 9.48-57.2). specificity 66.7% (95% Cl 42.8-90.5), PPV 20% (95% Cl -0.24-40.2) and NPV 80% (95% Cl 59.8-100.2).

Conclusion: Since the recent standard treatment for T3 and T4 tumors was to undergo preoperative chemoradiation, the lower accuracy occurred due to down-staging of the tumor from the neoadjuvant treatment. In addition, there was only a small number of included patients. which affected the statistical analysis of this study. The accurate results in the future study of transrectal ultrasound in preoperative staging of rectal cancer should be achieved by an increased sample size.